effective instruction for socially disadvantaged low-progress
TRANSCRIPT
Effective instruction for socially disadvantaged low-progress readers:
The Schoolwise Program*
Kevin Wheldall
MULTILIT Research Unit,
Macquarie University, Australia
This is a preprint of an article whose final and definitive form has been published in
the 'Australian Journal of Learning Difficulties' (c)2009 Copyright Taylor & Francis.
The Australian Journal of Leaning Difficulties is available online at:
http://www.informaworld.com
Wheldall, K. (2009). Effective instruction for socially disadvantaged low-progress
readers: The Schoolwise Program. Australian Journal of Learning Difficulties, 14 (2),
(in press).
* A spoken version of this article was originally to be presented by the author to mark
the presentation to him of the Mona Tobias Award for 2008 of Learning Difficulties
Australia. He was prevented from doing so as a result of illness.
Correspondence: Professor Kevin Wheldall; email: [email protected]
1
Abstract
In this article, I consider social class and reading performance, outline a non-
categorical approach to reading disability, describe the reading intervention program
we have developed for older low-progress readers, and seek to demonstrate how
students from socially disadvantaged backgrounds can, and do, make substantial
progress when offered effective reading instruction based on the available scientific
research evidence.
2
Effective instruction for socially disadvantaged low-progress readers:
The Schoolwise Program
In the fourth of his 2008 Boyer Lectures, Rupert Murdoch spoke frankly about
the failure of public education systems:
“The unvarnished truth is that in countries like Australia, Britain, and
particularly the United States, our public education systems are a
disgrace. Despite spending more and more money, our children seem
to be learning less and less – especially for those who are most
vulnerable in our society.” (Murdoch, 2008, p. 61).
He went on to say “The tragedy today is that in many nations like Australia,
the people who need a solid education to lift them out of deprived circumstances are
the people who are falling further and further behind” (p. 62). Deputy Prime Minister
and Education Minister, Julia Gillard, readily agreed with Murdoch the day following
his lecture, commenting that too many students were failing to meet minimum
standards and that most of these were from disadvantaged backgrounds. As will
become evident, these problems begin early, as a result of these students struggling to
learn to read in their first few years of schooling.
In this article, I shall describe the program with which I have been
involved since 1996, designed to meet the needs of older low-progress
readers from socially disadvantaged backgrounds. But, first, I would like to
provide a brief, more personal perspective on this topic.
A Personal Preamble
Since the Mona Tobias Award address is traditionally presented orally, I have
taken the liberty of providing a more informal account of my work in this area than the
usual, more formal research report and to reveal the roots of my interest in helping
socially disadvantaged low-progress readers. It should be made clear at the outset,
however, that the work on the Schoolwise Project reported here owes as much to my
project co-director, Dr Robyn Beaman, as it does to me. Having said this, I take
personal responsibility for the more general and critical views on education expressed
here and my interpretations of our research findings.
As a working class lad from a council housing (i.e. public housing) estate in
Derby in the UK, I made it to Manchester University just after the hippie ‘summer of
3
love’ in 1967, by the skin of my teeth. This was thanks to a pushy mother who had
made it her business to teach me to read at home when, as a result of moving school at
an inopportune time, I failed to do so in my first year or so of schooling. (She also
continually reminded me to ‘speak properly’, as she put it.) By dint of her efforts, I
had passed what was known as the ‘11 plus’ exam and had gained admission to a
quasi-grammar school. In those days, the doors were still open for bright working
class kids to escape the poverty trap. So, coming originally from a working class
background and also being English by birth, I have always been fascinated by social
class and the way it permeates almost every aspect of English (and, to a lesser extent,
Australian) society.
Partly as a result of taking subsidiary units in sociology in my first (honours)
degree in psychology (weekly rants from a Mancunian1 marxist in a second year unit
entitled ‘Class, Status and Power’ spring, unforgettably, to mind), but also because of
the hippie spirit of the times, I became a thoroughly radicalised ‘class warrior’. (This
was a far cry from the moderate who was subsequently selected as a parliamentary
candidate for the Social Democratic Party in the ‘Falklands’ general election in Britain
of 1983 - 10,613 votes but I came third, out of three!). Predictably, if paradoxically, I
also became thoroughly ‘embourgeoised’ in the process and learned from Nancy
Mitford, and later Jilly Cooper, just what was ‘U’ and what was ‘non-U’. (To this day,
I refer to serviettes as ‘paper napkins’ but I have forced myself to re-learn saying
‘toilet’ and to stop saying ‘the loo’ in Australia.) More seriously, I was also greatly
influenced by the seminal (if often misunderstood) work of the British sociologist,
Basil Bernstein, on ‘Class, Codes and Control’ (Bernstein, 1971) and his concept of
restricted and elaborated code users.
Having received a very patchy education in psychology (far too much Freud
and next to nothing on behaviour analysis), I was fortunate enough to be appointed,
immediately after graduation in 1970, as an academic research associate to Professor
Peter Mittler at his then newly formed Hester Adrian Research Centre (HARC) at
Manchester University, where I completed my doctoral research on receptive language
development (see Wheldall, 1976 for a summary). Peter and his colleagues patiently
filled in most of the gaps in my psychology education including a very thorough
grounding in research methodology and the work of B. F. Skinner on behaviour
1 Mancunian means a native of Manchester, UK.
4
analysis and its applications in special education. (I was still quite keen to ‘smash the
state’ at this point and Peter also deserves credit for putting up with a hippie research
associate with hair well below his shoulders).
As part of a continuing research program on the development of the Sentence
Comprehension Test (Wheldall, 1987; Wheldall, Mittler, & Hobsbaum, 1979, 1987),
after I left Manchester for a lectureship at Birmingham University in 1973, I began to
explore the effects of social class on receptive language development and how best to
help pre-school children from disadvantaged backgrounds by means of pre-school
education (Wheldall, 1978; Wheldall & Martin, 1977; Wheldall, Anderton, Bott, &
Kingslake, 1982). In short, we confirmed that sentence comprehension and vocabulary
were both strongly associated with social class (socio-economic status as measured by
father’s occupation) and that pre-school interventions could have a powerful effect in
improving the receptive language skills of young children from less advantaged home
backgrounds. We also found (Wheldall et al., 1982) strong links between social class
and reading performance at six years old. Young children from less advantaged home
backgrounds came to school with less well developed receptive language skills (unless
they had received some form of effective pre-school intervention) and this class effect
also manifested itself in the form of low progress in learning to read, sowing the seeds
for later school failure and, hence, the perpetuation of the class system.
If ever there were any residual doubt about the influence of social background
on students’ reading and related skills, such doubt has been expunged by the
publication of the findings from the first NAPLAN (National Assessment Program
Literacy and Numeracy) in 2008 (Ministerial Council on Education, Employment,
Training & Youth Affairs, 2008). From this first national survey of the performance of
Australian school students in Years, 3, 5, 7 and 9, it is clear that children’s home
backgrounds, with socio-economic status estimated by both level of parental education
and parental occupation, are closely linked to literacy skill levels. Social class and
reading performance are undoubtedly closely related, with students from less
advantaged home backgrounds typically underperforming in comparison with their
more advantaged peers. Quick calculations of the effect sizes (Cohen’s delta) for
differences in mean reading performance of students falling into the top and bottom
categories (for both parental education and parental occupation) in Years 3, 5 and 7
show them all to be large (>0.8), averaging 0.94. In other words, there is very nearly a
standard deviation difference in performance levels. Moreover, at least six times as
5
many students from the lowest category compared with the top category fall into Band
1, ‘Below National Minimum Standard’.
In this article, I shall describe the reading intervention we have developed for
older low-progress readers (MULTILIT) and seek to demonstrate how students from
socially disadvantaged backgrounds can, and do, make substantial progress when
offered effective reading instruction based on the available scientific research
evidence. Prior to this, however, it is important to spell out our overall approach to
teaching low-progress readers to provide a context for what follows.
A Non-Categorical Approach to Teaching Low-Progress Readers
Students struggle to learn to read for a variety of reasons. The actual causes of
reading disability are frequently not known, are more often hypothesised than proven,
and tend to be inferred from the fact that such students are failing to progress at the
same rate as their peers. This is one reason why I prefer to use the more neutral term
‘low-progress readers’.
As a result of scientific research carried out over the past 40 years, most
reading scientists now subscribe to variations of the phonological-deficit theory of
reading difficulties (reviewed Pogorzelski & Wheldall, 2005), that is that reading
problems are largely the result of language difficulties, specifically the ability to
segment and to blend the component sounds within words. Problems in phonological
processing may arise from intrinsic (biological) factors, extrinsic (environmental)
factors, or both. I have suggested a simple two factor model (Pogorzelski & Wheldall,
2005) to account for the likely outcomes of the interplay of these two main factors:
intrinsic, possibly heritable (Bishop, 2001) phonological processing difficulties and
extrinsic difficulties resulting from an impoverished language and literacy learning
environment. It is important to emphasise that intrinsic and extrinsic factors are
probably both differentially distributed (let us assume normally, in the absence of
evidence to the contrary). The practical import of this is that even students whose
inherent phonological processing ability is somewhat reduced as a result of intrinsic
(biological) factors, but whose performance is extraordinarily low because they have
suffered the ‘double whammy’ of this being coupled with a poor literacy learning
environment (at home and/or school), can relatively easily be helped once effective
instruction is provided. On the other hand, students who have enjoyed an optimal
language and literacy learning environment, but whose intrinsic phonological
6
processing ability is severely compromised, are probably still going to struggle even
when provided with exemplary, evidence-based best practice remedial instruction, and
are likely to need continuing support over many years. Friend, DeFries, and Olson
(2008) have recently provided empirical support for this on the basis of their study
involving 545 identical and fraternal twins of which at least one child from each set of
twins had a reading disability. They concluded that: “Genetic influence was higher and
environmental influence was lower among children whose parents had a higher level
of education compared with children whose parents had a lower level of education” (p.
1124) and that their results are supportive of the view that “poor instruction or lack of
reading practice may often be the main cause of reading disability in children from
low-SES families” (p. 1129). Theoretically, this is all very interesting but does it
actually help us much in practice?
Diagnosis
Endemic within the field of research and practice in reading disability has been
a preoccupation with the diagnosis of the underlying causes of reading difficulties,
specifically to identify those students who may be said to be dyslexic as against being
so-called ‘garden variety’ low-progress readers. In some countries, a formal diagnosis
of a learning disability such as dyslexia may make access more likely to increased
facilities and support, curricular and assessment dispensations, and, not least, special
funding. Until recent years, the most widely accepted definition and diagnosis of
dyslexia was predicated upon an observed discrepancy between a child’s reading
performance and his/her more general intellectual and/or verbal ability. This
‘discrepancy’ model has come under increasing attack over the past decade or so (see
Limbrick, Wheldall, & Madelaine, 2008 for a recent review) and is currently being
largely replaced, in the US and the UK at least, by a Response to Intervention (RtI)
model (described below).
But there is a more fundamental objection to this preoccupation with diagnosis
and categorisation, regardless of how or whether the categories may be substantiated
objectively. While the label of dyslexia, or learning disability more generally, may
afford some comfort to students struggling to read and to their parents, in our present
state of knowledge, at least, there are few or no implications of such a diagnosis for
specific instructional intervention that is any different from what we might offer any
student struggling to learn to read. Some of the loose talk surrounding diagnosis and
remediation of dyslexia presupposes a magic bullet to fix the reading disability
7
specifically of children with dyslexia, access to which they will somehow be denied
without an appropriate diagnosis. But the form of instruction we should offer should
be no different from that which we would offer to any struggling reader and this form
of instruction should be predicated upon evidence-based best practice as identified by
scientific research into effective instruction in reading and related skills.
While it is commonly believed that knowing the ‘cause’ will necessarily help
with the solution, this is not necessarily always the case in education. We do not
usually know why a student cannot read. Even if we have a very good idea, it still does
not help us. What can be even more pernicious is the belief that different causes will
necessarily need different treatments or teaching methods. The categories commonly
employed to describe students with apparently similar disabilities and difficulties are
sometimes questionable but, more importantly, are of little or no use in determining
the appropriate way to teach a particular student. How you teach children to read has
almost nothing to do with the clinical diagnosis and should be determined not by the
nature of the child's disabling condition but by a needs-based appraisal of the student's
current level of functioning. Although it might seem like useful information to know
the precise minutiae of why a person has a certain disability, very rarely does this
actually help us to teach them any better. In spite of considerable research into so
called ‘aptitude-treatment interactions’ (Howell, 1995), there is no convincing body of
evidence to suggest that students with different disabling conditions need different
forms of instruction; rather, it appears that effective instruction is effective instruction
generally.
It is for this reason that many scientifically orientated reading researchers and
practitioners now subscribe to what is known as a non-categorical approach to
teaching students with learning difficulties in the area of literacy (Wheldall, 1994;
Wheldall & Beaman, 2000; Wheldall & Carter, 1996). Quite simply, knowing whether
a child is dyslexic or not, or the reason why she or he has struggled to learn to read,
offers no help in determining what to do to help him or her to master the skills
necessary for reading and spelling. Consequently, it makes more sense to address the
problem of poor reading directly, regardless of the hypothesised causation, using the
most powerful instructional interventions that have been shown to be effective.
So what does this mean for practice? Quite simply, it requires us to focus on
the solution not the cause. A careful needs-based appraisal of the student’s current
level of performance in a skill area will help us to determine the entry point for
8
instruction and to decide on the method or program to be employed. This is not
necessarily to say that ‘one size fits all’; there may be several alternative instructional
methods or programs that have been shown to be effective but the selection of the
most appropriate program will be determined by the idiosyncratic needs of the
individual child and his or her responsiveness to instruction, not the category of his or
her disabling condition. I like to regard this as a truly child-centred approach to
education. A non-categorical approach also aligns very well with the Response to
Intervention model, providing a more sound basis for reaching conclusions about the
severity of a child’s difficulties in mastering reading and spelling skills than the
discrepancy models used in the past, as we shall see.
Response to Intervention
Response to Intervention (RtI) is a tiered model of instruction for students
experiencing difficulties in acquiring basic skills and appropriate social behaviours
(Bender & Shores, 2007; Council for Exceptional Children, 2008; Fuchs & Fuchs,
2007). Tiered instruction commonly (but not invariably) comprises instruction at three
increasing levels of intensity. In the context of literacy instruction, it is predicated
upon exemplary initial instruction in reading and related skills being provided at the
whole class level during the first 6 to 12 months of schooling. This is known as Tier
One instruction. Experiencing initial instruction based on evidence-based best practice
will ensure that the vast majority of students will get off to a good start in learning to
read and spell. Those students who begin to fall behind, often operationally defined as
those in the bottom 25% of what might be expected for the age cohort, are then offered
Tier Two instruction.
Tier Two instruction typically takes the form of more intensive, more targeted
small group literacy instruction, again based on what scientific research has shown to
be the most effective methods and curriculum content for teaching lower-progress
readers. Students are taught in small groups of four to six students, preferably by a
teacher or paraprofessional who is well versed and skilled in the delivery of effective
remedial instructional programs. Such instruction should be provided daily, if possible
for at least half an hour. This more intensive (and, of necessity, more expensive)
option is reserved only for those ‘failing to thrive’ under the regular classroom regime
of Tier One. Tier Two level intervention is likely to resolve the difficulties
experienced by the great majority of lower-progress readers and will enable them to
get ‘back on track’ and progressing at a similar level to their classroom peers. There
9
will always be a few students, however, who fail to respond even when offered this
more intensive level of Tier Two instruction and these students need Tier Three
intervention.
Tier Three intervention does not necessarily involve appreciably different
instruction from that offered in Tier Two except insofar as the instruction provided is
even more intensive, in more specifically targeted form, tailored to the specific needs
of the individual student on a one-to-one basis, and preferably provided by a reading
expert.
Within the RtI model, students with a learning disability such as dyslexia are
defined as those students needing Tier Three intervention, students who are still
struggling even when they have been offered both exemplary initial reading
instruction (Tier One) and subsequent exemplary remedial instruction (Tier Two).
These are typically the students who are likely to need continuing literacy support,
possibly over many years. In our present state of knowledge, we have no way of
telling in advance just who these students will be other than by using the ‘suck it and
see’ approach of RtI coupled with very careful progress monitoring. The question then
arises of how best to judge whether a child is responding well to instruction or not and
hence whether she or he needs Tier Two or Tier Three intervention. Traditional
reading tests are of limited use for this purpose since they are not usually sensitive
enough to pick up small gains over short periods of time, nor should they be repeated
frequently or after only a short time interval, if they are to provide reliable measures.
On the other hand, it would not be in the best interests of the child to remain for too
long on a program under which he was failing to progress and which will only be
detected when he is retested at the end of the year or even after six months.
Progress Monitoring Using Curriculum-Based Measurement (CBM)
Curriculum-based fluency measures have emerged as the preferred alternative
for progress monitoring (Fuchs, 2004; Madelaine & Wheldall, 1999, 2004). Oral
reading fluency may be measured by passage reading tests, having a student read
aloud a passage of text and counting the number of words read correctly in one
minute. This seemingly rather crude index in fact correlates very highly with other
more complex measures of reading including both reading accuracy and reading
comprehension. Because this exercise may be repeated frequently with different
samples of text from the curriculum at a similar level of difficulty, this provides a
10
ready means of tracking progress over time. Moreover, these curriculum-based
measures are quick and easy to administer.
Curriculum-based measurement provides the means by which we can
determine the tier of support a student needs. First, a passage reading test (or tests)
may be used as a simple screening instrument to determine which students are
struggling to keep up with their peers, say the bottom 25%. These students are then
offered Tier Two small group remedial instruction. By subsequently monitoring
progress on a weekly or even fortnightly basis over about six weeks we may then
determine who is responding readily and who is not and hence who is likely to need
more intensive, individualised Tier Three instruction for greater and continuing
duration, and even whom we might choose to refer to as having ‘dyslexia’, for
administrative/funding purposes.
Thus it may be seen how a non-categorical approach to remedial instruction for
low-progress readers aligns very well with the Response to Intervention model,
supported by curriculum-based measurement, and offers the best option for helping
low-progress readers in our present state of knowledge. This non-categorical approach
to teaching low-progress readers has defined our own work in this area.
A Brief Description of the MULTILIT Reading Tutor Program
In an earlier article in this journal (Ellis, Wheldall, & Beaman, 2007), we
described the research locus and conceptual basis for our approach to remedial
reading instruction, Making Up Lost Time In Literacy (or MULTILIT), and so I will
not dwell on this in detail again here (see also Wheldall & Beaman, in press). In our
view, and on the basis of the available scientific evidence, the most effective remedial
programs for low-progress readers take a scientifically balanced perspective and
involve intensive, systematic instruction in three main areas: phonic word attack
skills; sight word recognition; and supported book reading in a one-to-one context.
The (revised) MULTILIT Reading Tutor Program (RTP) (MULTILIT, 2007a)
incorporates all three of these key features and forms the core of what we offer to
low-progress students. The program was specifically designed for teaching low-
progress readers in Year 2 and above who are reading at a level considerably below
what might be expected for their age and who have not acquired the basic skills
needed to become functional readers.
11
MULTILIT Word Attack Skills
Children learning to read primarily need to learn how to ‘crack the code’ -
how to decode words they have not previously encountered by breaking words down
into their component phonic parts. Low-progress readers need intensive, systematic
instruction both in how to break up (‘segment’) words into their component letter
sounds and, even more importantly, how to ‘blend’ component letter sounds into
words. MULTILIT Word Attack Skills is designed to do precisely this: to teach low-
progress readers the phonic word attack skills essential for rapid decoding, using a
synthetic phonics approach (MULTILIT, 2007b).
MULTILIT Sight Words
Sight words are words that can be read automatically on sight without
recourse to decoding strategies. When learning to read, it makes good sense for low-
progress readers to learn a small corpus of very common sight words so that they will
not need to struggle to decode every single word that they encounter in a sentence.
MULTILIT Sight Words (MULTILIT, 2007c) systematically teaches the automatic
recognition of 200 high frequency sight words. The 200 words included were
empirically determined from the most frequently occurring words encountered in
children’s books (excluding proper nouns and simple, easily decoded, consonant-
vowel-consonant or CVC words) on the basis of the research carried out by Morag
Stuart and her colleagues in the UK (Stuart, Dixon, Masterson, & Gray, 2003 a, b, c).
MULTILIT Reinforced Reading
Reinforced Reading (Wheldall & Beaman, 2007) is a program developed to
enhance the student’s independent reading skills, to increase reading fluency, to build
vocabulary and to foster comprehension. It constitutes a re-engineered version of the
well-known set of tutoring strategies for use with low-progress readers known as
Pause, Prompt and Praise (PPP) (Glynn, 1987; McNaughton, Glynn, & Robinson,
1981; Wheldall & Mettem, 1985), in which the role of phonics has been emphasised
and the reliance on contextual clues has been relegated to a reader self-checking role.
The aim of the tutoring session is for the tutor to listen to the low-progress reader read
natural language books at an appropriate level of difficulty for up to 15 minutes,
following an introduction by the tutor. The tutor is trained to pause for up to 5
seconds or wait until the end of a sentence when a mistake is made to permit time for
self-correction. If no self-correction occurs, the tutor supplies up to two prompts in
the form of a general phonic prompt (“How does this word begin?”, “What sound do
12
these letters make?”) or a re-read prompt (“Read that again from the beginning of the
sentence.”), followed by a specific phonic prompt if the first prompt is not successful.
When the student correctly reads a sentence or paragraph, self-corrects without a
prompt, or successfully uses a given prompt to identify a word, specific praise is
given. At the end of the session the tutor asks questions about the passage just read. In
addition, two variants of this basic procedure target fluency and comprehension
specifically.
These MULTILIT one-to-one teaching activities provide sequential learning
for students who are behind in reading skills. The MULTILIT Program is typically
implemented by a teacher but a teacher’s aide, trained volunteer, trained parent, or
skilled peer tutor, working under the direction of a teacher, could also implement the
program. Each child is tested on entry to the MULTILIT Program to place them at the
appropriate levels in each component of the MULTILIT Reading Tutor Program. The
lessons in this program concentrate on decoding skills (MULTILIT Word Attack
Skills) and accurate and automatic recognition of sight words (MULTILIT Sight
Words), and the practice and generalisation of these skills using the connected reading
of real text in MULTILIT Reinforced Reading.
In this article, I want to discuss the findings from a specific application of
MULTILIT for socially disadvantaged students, the Schoolwise Program.
The Schoolwise Program
In 1994, the Rev Bill Crews, chairman of the Exodus Foundation (a charity
focusing on the plight of the homeless), first approached me with his concerns
regarding ‘streetkids’ and disaffected youth and his desire to implement a preventative
program to keep students in school and off the streets. For many of these students, the
problems begin early as a result of initial academic failure in learning basic skills and
are then exacerbated by the increasing demands made by a largely text-based
curriculum predicated upon mastery of the very skills in which they are most deficient.
Schooling can become an increasingly aversive experience for many such
marginalised students.
Since 1996, the Exodus Foundation has funded a research consultancy
annually for the MULTILIT team to run a version of the MULTILIT program, known
as the ‘Schoolwise Program’, on the Exodus Foundation site at Ashfield in Sydney.
The aim of the project has been to address the needs of older low-progress readers
13
from socially disadvantaged backgrounds and who were at risk of serious disaffection
from school. The success of the Schoolwise Program in teaching successive intakes of
older low-progress readers to learn reading and related skills has been amply
documented in our commissioned research report to the Australian Commonwealth
Government Department of Education, Training and Youth Affairs (DETYA), ‘An
Evaluation of MULTILIT’ (Wheldall & Beaman, 2000,) which detailed the efficacy of
MULTILIT.
In essence, the Schoolwise Project aims to redress substantially the literacy
skills deficits of older low-progress 'at risk' students in their final years of primary
school (or first year of high school) by providing a program of intensive, systematic,
skills-based literacy instruction. Students typically attend the Schoolwise Program
every morning for two school terms in a special learning tutorial unit (the Exodus
Foundation Tutorial Centre) based in a community centre near to, but not directly
associated with, local schools. Thus the program also provides respite within a
positively focused learning environment that is clearly and deliberately differentiated
from 'school'.
Students are referred by their schools on the basis of social disadvantage, their
low levels of literacy (as measured by the Neale Analysis of Reading), their inability
to cope with the regular curriculum, and the substantial risk of their disaffection from
school. Our own assessment procedures (curriculum-based) prior to student selection
provide confirmation of the student’s suitability and need.
Research findings based on the results from, and experience with, successive
intakes into the Schoolwise Program since the first intake in 1996 have consistently
testified to the success of this literacy intervention program (Wheldall & Beaman,
2000). In less than 5 months of instruction, successive intakes have made very large
average gains in reading accuracy, reading comprehension, single word recognition,
reading fluency and spelling. These were students who had made little or no progress
in recent years and who were typically 3 or more years behind in reading and related
skills when they entered the program.
The commissioned report to DETYA ‘An Evaluation of MULTILIT’
(Wheldall & Beaman, 2000) (referred to above) included the findings for the first three
years of the Schoolwise Program, 1996 to 1998. In those years, MULTILIT was
delivered individually to students as a one to one program; what we would regard
today as essentially a Tier Three intervention. Over the past five years or so, however,
14
we have developed a small group instruction version of the program suitable for use as
a Tier Two intervention.
The Intervention
Over time the Schoolwsie Program delivered has changed in both length of
daily intervention and type. From the late 1990s until the end of 2003 the program was
delivered between 8:45am to 1pm daily including a brief (15 minute) break, resulting
in four hours of instructional time daily. The mode of the core MULTILIT Program
delivery was on a 1:1 basis, with the instructor working with each child in his/her
group successively. (Some of the other lessons during the morning program were
group-based, however e.g. SRA Spelling Mastery.)
Over the course of 2004, the main mode of delivery for the MULTILIT core
elements changed to a group instruction format and the instructional time was reduced
from 4 hours (plus a break) to three hours (inclusive of a 15 minute break). We had
effectively reduced the instructional time to deliver the program from 4 hours to 2
hours 45 minutes. This change enabled students to get back to school earlier in the
school day.
The Schoolwise Program at the Exodus Foundation Tutorial Centre provides
intensive MULTILIT literacy instruction, five mornings per week, for two school
terms (approximately 20 weeks) for each cohort of 36 students. As described earlier,
the core MULTILIT Program consists of MULTILIT Word Attack Skills, MULTILIT
Sight Words and MULTILIT Reinforced Reading (using Pause, Prompt and Praise),
and also utilises other supportive evidence–based programs.
Students (since Semester 2, 2004) attend the program from 8:30am to 11:30am,
Monday to Friday for two school terms. The daily MULTILIT session consists of: 25
minutes of group MULTILIT Word Attack Skills (two separate lessons for accuracy
and fluency); 15 minutes of group MULTILIT Sight Words; 25 minutes of group
MULTILIT Reinforced Reading; 20 minutes of group spelling (using the Spelling
Mastery program, a direct instruction program from SRA); one hour of ‘home group’
where individual sessions and independent work are completed, and 20 minutes of
peer-tutoring using Reinforced Reading.
The development of a version of the MULTILIT program for delivery in small
groups has led to an increase in the cost-effectiveness of the program, delivering the
intervention to a larger number of students in a shorter time period. This has also
meant a considerable reduction in the amount of time students are out of their regular
15
school environment each day, with students completing their daily program by
11.30am rather than 1pm, as previously mentioned.
One of the key features of a group-based MULTILIT intervention (as opposed
to an individual program of MULTILIT instruction) is that students are placed in
groups at the correct level in terms of their ability. Group composition changes for
each program component and this is a dynamic process. In addition, the collection and
use of program data to inform instructional decision-making is a key feature, with data
being reviewed weekly to determine group composition and to monitor progress.
In addition to MULTILIT, students also receive the SRA Spelling Mastery
program (Dixon, Engelmann, & Bauer, 1999). This is a rules-based direct instruction
program with six levels designed to provide increasingly independent practice of the
skills taught through to mastery. It is used in conjunction with the MULTILIT
Program, as it is complementary in its treatment of letter-sound knowledge. A
placement test is used to determine a student’s entry level.
Activities during the home-group period typically include a one-to-one session
with the instructor, a one-to-one Reinforced Reading session with a community
volunteer, computer program activities and independent work. Each student has an
Independent Folder that contains work to be completed during the period, as per their
work contract. The work set for the student’s Independent Folder is at an independent
level in terms of difficulty. Students are able to revise, practise and generalise their
skills through worksheets and journal writing activities while learning to work
independently. Instructors mark the folder each day and students are responsible for
correcting their mistakes the following day. Students are rewarded (using a system of
points redeemable for small trinkets) for completing all of the assigned work in their
work contract each week.
Developing independent work habits is an important aspect of the program, as
many older low-progress students become dependent learners as a result of their
learning difficulties. Ensuring that the work to be completed is at an appropriate and
achievable level is an important antecedent to independent work completion. During
this home group session, students also receive one-to-one instruction in MULTILIT
Word Attack Skills (accuracy and fluency) and MULTILIT Sight Words, if needed.
The need for and the amount of individual sessions a student has with an instructor is
determined by the result of individual testing that is conducted each week during the
one-to-one session. In this way, the students with the greatest needs are provided with
16
more intensive individual instruction. Students are mainly taught in small groups (Tier
Two) but are offered more intensive instruction (Tier Three) if they are shown to be
struggling to make good progress.
As already mentioned, students also get the opportunity of reading with a
volunteer community tutor. The volunteer tutors are trained and monitored in the
delivery of the MULTILIT Reinforced Reading strategies using the revised and
updated version of Pause, Prompt and Praise, as described earlier. This individual
attention with a supportive ‘other’ is a very important element of the program,
ensuring daily practice using instructional level text (90-95% accuracy) for at least 20
minutes. A focus on reading comprehension is included in this session and students are
not only required to utilise their word attack skills and sight word knowledge, but must
be able to answer questions, recall and recap the story or information in the text. This
daily session provides opportunity for the generalisation of the skills directly taught in
the program in a more naturalistic text–reading context.
The Schoolwise Program also employs a range of complementary programs in
order for students to be able to generalise their skills. For example, once students have
completed the MULTILIT Sight Words and/or MULTILIT Word Attack Skills
programs, they may move on to comprehension or writing programs, such as the SRA
Reasoning and Writing program (Engelmann & Davis, 2001).
An Evaluation of Efficacy
For the purposes of this brief report of our more recent findings, I have
combined the data from the last two intakes of students from Semester 1 and Semester
2 of 2008 who participated in and completed the Schoolwise Program for a full two
terms. Over the year, 67 students completed the Schoolwise Program and were present
for the pre- and post-program assessments. All students were from Years 5 (40) and 6
(27); 40 were boys and 27 were girls. The mean age of the students on program entry
was 11 years (132 months; ranging from 115 to 148 months). Each intake of students
completed two terms of instruction.
At the commencement of each program, students are typically given a battery
of standardised tests of reading and related skills (‘the MULTILIT Battery’),
administered by trained research assistants. The battery consists of measures of
reading accuracy and comprehension (Neale Analysis of Reading; Neale, 1999), single
word recognition (Burt Word Reading Test; Gilmore, Croft, & Reid, 1981), spelling
(South Australian Spelling Test; Westwood, 1999), oral reading fluency using a
17
curriculum-based measure (the Wheldall Assessment of Reading Passages; Wheldall,
1996; Wheldall & Madelaine, 2006) and phonological recoding using a non-word
reading test (Martin and Pratt Nonword Reading Test; Martin & Pratt, 2001). Students
are tested again on the entire battery at the end of the typical two-term (20 weeks)
program.
At program commencement, the crude average reading age for reading
accuracy for these students as measured by the Neale Analysis was 93 months (7 years
and 9 months), more than 3 years below chronological age. In terms of Neale reading
comprehension, the crude average reading age was 88 months (7 years and 4 months),
again more than 3 years below chronological age. The verbal ability of the students at
the beginning of the study, as estimated by the Peabody Picture Vocabulary Test
(Dunn & Dunn, 1997), averaged 90.27 (SD 15.34) (standardized score).
The results for these students attending the Schoolwise Program, showing pre-
test and post-test means and standard deviations for all measures (raw scores), after 18
weeks of instruction, are shown in Table 1. In under 5 months of participation in the
Program, these students made crude average gains of 20 months in Neale reading
accuracy, 16 months in Neale reading comprehension, 20 months in Burt single word
reading, 22 months in spelling, 30 months on the non-word test of phonological
recoding and could read 46% more words correctly per minute on the WARP (reading
fluency).
TABLE 1 ABOUT HERE
Analyses of raw scores indicated statistically significant gains (p<0.0005) on
all of these measures. In order to appreciate the magnitude of these gains, we may also
examine the relevant effect sizes (ES). The effect sizes (Cohen’s delta) were all large
(>0.8) ranging from 1.18 to 1.49. These results, then, are shown to be of high
educational significance, as well as statistical significance. The performance of the
first intake as a whole on a weekly basis over the two terms of instruction is shown in
Figure 1. Reading fluency, as measured by performance on the WARP weekly
progress monitoring passages, in terms of words read correctly per minute (wcpm), is
seen to rise steadily and substantially over the period, from a mean of 76 wcpm to 111
wcpm.
FIGURE 1 ABOUT HERE
18
A Waitlist Control Study of the Efficacy of the Schoolwise Program
In recent years we have witnessed increasing demand for evidence-based
practice in education, a very welcome development, in my view. Unfortunately, as
Carter and Wheldall (2008) have argued, evidence of sufficient rigour and quality
upon which to base educational decisions is thin on the ground. While there is a
plethora of educational research, much, if not most, educational research these days
tends to be qualitative or otherwise non-empirically based (Wheldall, 2006). Very few
so-called ‘gold-standard’ truly experimental studies of programs, interventions and
innovations are conducted and, it must be admitted, MULTILIT is no exception.
While we now have a very large database testifying to the efficacy of the MULTILIT
Schoolwise program compiled over the years since 1996 and comprising data on
nearly a thousand low-progress readers, we, like many program developers, have
experienced difficulties in completing true randomised control trials.
These data, although not truly experimental since we were not able to employ
control groups, nevertheless do withstand reasonably close scrutiny with regard to
evidence for efficacy.
• First, the gains made have been consistently substantial, statistically
significant, replicated over many years now, and the effect sizes based on gains
made between pre- and post assessments have generally been large or very
large.
• Second, while we do not have actual control group data from randomly
assigned groups, we have analysed data from similar samples of low-progress
readers over similar time periods who did not receive MULTILIT interventions
and gains of this magnitude were not apparent in these samples (Wheldall &
Beaman, 2000). We concluded that the typical rate of progress for older low-
progress readers from Year 5 was typically very low; not surprisingly since to
be defined as a low-progress reader (i.e. over 2 years behind age peers and in
the bottom 25% of the age cohort), students could logically only have been
making very low progress.
• Third, while other factors could explain the results (e.g., novelty and the
experience of being selected for a special program), our findings are unlikely
merely to be reflections of the regression to the mean effect. The students were
nominated for the program by their home schools on the basis of need and low
performance on the Neale Analysis administered by the school using a
19
different parallel form of the Neale to the one in our battery and administered
prior to our pre-testing. They were not assigned to the program on the basis of
our pre-testing. Thus, any regression effects resulting in increased score at
retest, as a result of the less than perfect reliability of the measures, should
have been evident in our own subsequent pre-test scores on the Neale and
thereby would reduce the size of gain between our pre- and post-testing. In
other words, our results probably already take possible regression effects into
account.
• Fourth, as we shall see, weekly graphing of the progress of the intake into the
program using the curriculum-based measure, the WARP test, show that
reading performance rises steadily over the two terms. This is no trivial point
since we can effectively show the learning taking place regularly over time;
further evidence that this is not a simple regression effect.
During 2007, however, the opportunity arose of conducting a quasi-
experimental, waitlist control study (Wheldall & Beaman, in preparation for
publication). The small group version of the Schoolwise program was delivered to 36
socially disadvantaged Year 5 and Year 6 students in each of the two semesters in
2007. While not randomly allocated to conditions they comprised students from the
same overall pool of Schoolwise applicants attending local schools. All students were
assessed on the MULTILIT battery at the beginning of the first semester, at the end of
the first semester and again at the end of the second semester. During the first
semester, the first experimental group of students attended the Schoolwise program
daily for 3 hours each morning at the Exodus Tutorial Centre while the waitlist control
students continued with their regular curriculum in their home schools. In the second
semester, the experimental group students who had been receiving MULTILIT every
morning returned to their regular classes full-time and the former waitlist control
students became the second experimental group receiving MULTILIT every morning
in the tutorial centre, in a cross-over design. We obtained complete data sets on 35 of
the 36 students in each group, 70 students in total.
At the commencement of the study, students in both groups were the same age
(about 10 years 10 months) on average and both groups comprised very similar ratios
of boys to girls and Year 5 to Year 6 students. The verbal ability of the two groups at
the beginning of the study, as estimated by the Peabody Picture Vocabulary Test
(Dunn & Dunn, 1997) was also very similar. The mean raw scores were slightly higher
20
for the waitlist control group on three of the six measures comprising the MULTILIT
Test Battery while they were slightly lower on the other three tests but none of these
differences were statistically significant.
After the first semester, the experimental group made far greater gains than the
waitlist control group on all measures. Analyses of covariance of mid-year test raw
scores (with initial pre-test scores as a covariate) revealed that all differences in gain
between the two groups at mid-year testing were statistically significant (p<0.001,
reading comprehension, p<0.01). The following table (Table 2) shows the effect sizes
comparing the mean raw scores of the two groups at mid-year testing and also the
effect sizes for the original experimental group comparing pre and post intervention
mean raw scores. As may be seen, all effect sizes, were at least moderate in size (i.e. >
0.5) and mainly large (i.e.> 0.8), demonstrating strong effects.
TABLE 2 ABOUT HERE
As further confirmation, we may also examine the performance of the control
group after cross-over, when it became the second experimental group. In brief, these
former control students subsequently also made major gains on the measures included
in the test battery following two terms in the Schoolwise Program, with large effect
sizes evident for all measures (0.88 to 1.42). Consequently, we may claim that our
quasi-experimental waitlist control study has provided more rigorous evidence for the
efficacy of the Schoolwise MULTILIT program, but a true ‘gold standard’ randomized
control group study has yet to be completed. This should, preferably, be undertaken by
an independent, external research group. One of the problems of providing this much-
needed experimental data, however, is that we (like all program developers) are
dependent upon independent research groups choosing to evaluate our program and
having the resources available to do so. Clearly, if we commissioned such a study we
would still be compromised. This problem has very broad implications in our quest for
truly evidence-based practice generally.
So Where Do We Go From Here?
In this paper, I have endeavoured to demonstrate that low-progress readers from
socially disadvantaged backgrounds can and do make substantial gains in reading
performance following scientific evidence-based programs of literacy instruction such as
MULTILIT. The question now is how do we deliver such instruction to as many
struggling low-progress readers as we can, in a cost effective way?
21
First, at the risk of preaching heresy, I do not believe that we need necessarily to
spend more money. The problem of education generally is not too little funding; it is how
it is spent. The Auditor General of New South Wales drew attention to this in a recent
report:
“Since 1998-99 funding for literacy and numeracy programs has increased
three-fold from $53 million to $154 million in 2006-07. … Despite this, over
the last decade State tests have shown little change in results for numeracy
and literacy, both in terms of the percentages of students in the performance
bands and the state average scores.” (Audit Office of New South Wales,
2008, p. 3).
Secondly, nor do I believe that is it because we have too few teachers. We do not
need more teachers; perhaps we need fewer. … We might be better served if we spent
part of the education salary budget on fewer, better paid, far better trained educators of
higher initial academic quality, and spent the remainder on paraprofessionals. The
profession of teacher has become increasingly devalued by the community and I propose
for consideration instead the new profession of educator. We should be seeking our
brightest and our best to enter this profession and pay them accordingly. Educators would
be carefully selected, academically very bright and very competent to begin with, and
highly educated and trained to at least Masters degree level in the science of effective
instruction and classroom management as well as in specific curricular areas. They would
function more as managers of instruction responsible for designing programs and
managing the academic progress of the equivalent, perhaps, of two regular classes and
leading a team of educators in training and paraprofessionals who would carry out the
day to day instruction under the educator’s guidance.
Very few of the MULTILIT tutors who have delivered the Schoolwise Program
over the years have been qualified teachers. They have typically been psychology
graduates or undergraduates or trainee special educators from non-teaching backgrounds
employed as instructors and tutors to deliver the program and to employ Positive
Teaching techniques. They typically do not require extensive training because the
programs are clear and systematic and the skills are relatively easily mastered. This is not
to say that anyone can deliver MULTILIT; some are temperamentally, or even sometimes
philosophically, unable to deliver the program to the standard required. Trained teachers
sometimes even take a little longer to unlearn some of the misconceptions about
instruction and how reading works that they unfortunately learned as part of their
22
professional education at university or from within misguided state education
departments.
Having said this about paraprofessionals delivering the program, the programs
and teams of tutors must be led by high-level educators or special educators as defined
above. These educators make the instructional decisions regarding the program for each
child, collect and analyse independent progress data and make programming decisions
based on these data (such as promotion to the next book level, for example), mentor the
tutors and, generally, take direct responsibility for the progress of all students under their
remit.
And what of the future for MULTILIT? The MULTILIT Research Unit at
Macquarie University is currently researching and developing two new intervention
programs for younger children. The first ‘MINILIT’, or Meeting Initial Needs In
Literacy, is a small group-based program for students who have failed to make adequate
progress in reading during their first year of schooling. The aim is to provide both a more
effective and a more cost-effective alternative to the ubiquitous Reading Recovery for
young low-progress readers since Reading Recovery is only moderately effective
generally, is less effective with students with more serious phonological processing skill
deficits, and is very expensive (Reynolds & Wheldall, 2007; Reynolds, Wheldall, &
Madelaine, 2009. The findings from our preliminary studies have been very encouraging
(Reynolds, Wheldall, & Madelaine, 2007a, 2007b).
The second new program is for even younger children. Currently known as
PRELIT, it targets young children in their final year of pre-school prior to beginning
formal schooling in Kindergarten/Prep/Reception. It does not seek to teach young
children to read but rather seeks to ensure that children begin school with the necessary
prerequisite skills to learn to read. Consequently, PRELIT currently has two major foci:
structured instruction in phonemic awareness, emphasising both segmenting and blending
the sounds in words; and structured story book reading to enhance vocabulary knowledge
and to build listening comprehension skills. PRELIT is likely to be particularly necessary
for those children who do not typically enjoy the sort of language and literacy learning
experiences that most children from advantaged home backgrounds take for granted.
So, in a sense, I have come full circle, returning to my earliest research interests in
receptive language and the need to provide effective pre-school education for young
children from socially disadvantaged backgrounds. It is only when all children come to
school in roughly the same position on the starting grid that we can hope to see almost all
23
children learn to read quickly and easily. And since literacy underpins everything in
terms of future success in school and beyond, it is our greatest hope for ensuring a ‘fair
go’ for all Australians regardless of their social background.
24
Acknowledgements
I would like to acknowledge and record my appreciation of the warm support I have
received from the following individuals who have all helped to make possible the
research reported here:
• Dr Robyn Beaman, both my wife and my research partner, who has jointly
directed the Schoolwise Research Project with me since 1996 and who has
generously shared her expertise and ideas, as reflected in this paper;
• Rev Bill Crews and his colleagues at the Exodus Foundation, Ashfield, who
worked with us to establish the MULTILIT Schoolwise Program and have
funded its operation from 1996 to 2008;
• Dr Alison Madelaine and Georgia Callaghan, who have provided invaluable
consultancy support to the Schoolwise projects in recent years;
• And, not least, the MULTILIT Development Team and tutorial staff members
(far too numerous to mention individually) who have been involved as
MULTILIT instructors and tutors, research assistants, and project management
personnel.
25
References
Audit Office of New South Wales. (2008). Improving literacy and numeracy in NSW
public schools: Department of education and training. Sydney: The Audit
Office of New South Wales.
Bender, W. N., & Shores, C. (Eds.). (2007). Response to intervention: A practical
guide for every teacher. Thousand Oaks: Corwin Press.
Bernstein, B. (1971). Class, codes and control (volume 1). London: Routledge and
Kegan Paul.
Bishop, D. V. M. (2001). Genetic influences on language impairment and literacy
problems in children: Same or different? Journal of Child Psychology and
Psychiatry, 42, 189–198.
Carter, M., & Wheldall, K. (2008). Why can’t a teacher be more like a scientist?
Science, pseudoscience and the art of teaching. Australasian Journal of
Special Education, 32, 5-21.
Council for Exceptional Children. (2008). Responsiveness to Intervention: A collection
of articles from Teaching Exceptional Children. Washington, DC: Council
for Exceptional Children.
Dixon, R., Engelmann, S., & Bauer, M. (1999). Spelling mastery level b. SRA:
McGraw-Hill.
Dunn, L. M., & Dunn, L. M. (1997). PPVT-III: Peabody picture vocabulary test.
Circle Pines, Minn: American Guidance Service.
Ellis, L., Wheldall, K., & Beaman, R. (2007). The research locus and conceptual basis
for MULTILIT: why we do what we do. Australian Journal of Learning
Disabilities, 12, 61-65.
Engelmann, S., & Davis, K. L. S. (2001). Reasoning and writing level A: A direct
instruction program. Columbus, Ohio: SRA.
Friend, A., DeFries, J. C., & Olson, R. K. (2008). Parental education moderates
genetic influences on reading disability. Psychological Science, 19, 1124-
1130.
Fuchs, L. S. (2004). The past, present, and future of curriculum-based measurement
research. School Psychology Review, 33, 188-192.
Fuchs, L. S., & Fuchs, D. (2007). A model for implementing responsiveness to
intervention. Teaching Exceptional Children, 39, 14-20.
26
Gilmore, A., Croft, C., & Reid, N. (1981). Burt word reading test: New Zealand
revision. Wellington: New Zealand Council for Educational Research.
Glynn, T. (1987). More power to the parents: Behavioural approaches to remedial
reading at home. In K. Wheldall (Ed.), The behaviorist in the classroom.
London: Allen and Unwin.
Howell, K. (1995). Learning styles instruction: Questions and answers about aptitude
by treatment interactions. Special Education Perspectives, 4, 11-16.
Limbrick, L., Wheldall, K., & Madelaine, A. (2008). Gender ratios for reading
disability: Are there really more boys than girls who are low-progress
readers? Australian Journal of Learning Difficulties, 13, 161-179.
Madelaine, A., & Wheldall, K. (1999). Curriculum-based measurement of reading: A
critical review. International Journal of Disability, Development and
Education, 46, 71-85.
Madelaine, A., & Wheldall, K. (2004). Curriculum-based measurement of reading:
Recent Advances. International Journal of Disability, Development and
Education, 51, 57-82.
Martin, F., & Pratt, C. (2001). Martin and Pratt nonword reading test. Camberwell,
Victoria: Australian Council for Educational Research.
Ministerial Council on Education, Employment, Training and Youth Affairs. (2008).
2008 National Assessment Program Literacy and Numeracy: Achievement in
Reading, Writing, Language Conventions and Numeracy”. Canberra:
Ministerial Council on Education, Employment, Training and Youth Affairs.
McNaughton, S. S., Glynn, T., & Robinson, V. (1981). Parents as remedial reading
tutors: Issues for home and school. Wellington: New Zealand Council for
Educational Research.
MULTILIT. (2007a). The MULTILIT reading tutor program (revised). Sydney:
MULTILIT Pty Ltd.
MULTILIT. (2007b). MULTILIT word attack skills (revised): Manual. Sydney:
MULTILIT Pty Ltd.
MULTILIT. (2007c). MULTILIT sight words (revised): Manual. Sydney: MULTILIT
Pty Ltd.
Murdoch, R. (2008). A golden age of freedom. Sydney: ABC Books.
Neale, M. D. (1999). Neale analysis of reading ability – third edition. Hawthorn:
Australian Council for Educational Research.
27
Pogorzelski, S., & Wheldall, K. (2005). The importance of phonological processing
skills for older low-progress readers. Educational Psychology in Practice,
21, 1-22.
Reynolds, M., & Wheldall, K. (2007). Reading Recovery twenty years down the track:
Looking forward, looking back. International Journal of Disability,
Development and Education, 54, 199-223.
Reynolds, M., Wheldall, K., & Madelaine, A. (2009). The devil is in the detail
regarding the efficacy of Reading Recovery: A rejoinder to Schwartz,
Hobsbaum, Briggs, and Scull. International Journal of Disability,
Development and Education, 56, 17-35.
Reynolds, M., Wheldall, K., & Madelaine, A. (2007a). Developing a ramp to reading
for at-risk year one students: A preliminary pilot study. Special Education
Perspectives, 16, 39-69.
Reynolds, M., Wheldall, K., & Madelaine, A. (2007b). Meeting Initial Needs In
Literacy (MINILIT): Why we need it, how it works, and the results of pilot
studies. Paper presented to the Annual Conference of the Australian
Association of Special Education, Sydney, September 29, 2007. Australasian
Journal of Special Education, 31, 147-158.
Stuart, M., Dixon, M., Masterson, J., & Gray, B. (2003a). Children’s early reading
vocabulary: Description and word frequency lists. British Journal of
Educational Psychology, 73, 585-598.
Stuart, M., Dixon, M., Masterson, J., & Gray, B. (2003b). Children’s early reading
vocabulary: Description and word frequency lists. Retrieved June 15, 2007,
from http://www.ioe.ac.uk/phd/llrc
Stuart, M., Dixon, M., Masterson, J., & Gray, B. (2003c). Children’s printed word
database (Version 1.3, modified June 9, 2003). Retrieved June 15, 2007,
from http://www.essex.ac.uk/psychology/cpwd
Westwood, P. (1999). Spelling: approaches to teaching and assessment. Camberwell:
Australian Council for Educational Research Press.
Wheldall, K. (1976). Receptive language development in the mentally handicapped. In
Berry, P. (Ed.), Language and Communication in the Mentally Handicapped.
London: Arnold.
28
Wheldall, K. (1978). Social class and receptive language development in mentally
handicapped children. Bulletin of the College of Speech Therapists, No. 313,
5-7.
Wheldall, K. (1987). Assessing young children's receptive language development: a
revised edition of the Sentence Comprehension Test. Child Language
Teaching and Therapy, 3, 72-86.
Wheldall, K. (1994). Why do contemporary special educators favour a non-categorical
approach to teaching? Special Education Perspectives, 3, 45-47.
Wheldall, K. (1996). The Wheldall Assessment of Reading Passages: Experimental
edition. Sydney, Australia: Macquarie University Special Education Centre.
Wheldall, K. (2006). When will we ever learn? In K. Wheldall (Ed.), Developments in
educational psychology: How far have we come in 25 years? London:
Routledge
Wheldall, K., Anderton, P., Bott, E., & Kingslake, B. (1982). Socio-environmental
influences on the receptive language development of children starting school:
three follow-up studies after one year in school. Collected Original
Resources in Education, 6(1), Fiches 3D8-l3Bl2.
Wheldall, K., & Beaman, R. (2000). An evaluation of MULTILIT: ‘Making Up Lost
Time In Literacy’. Canberra: Department of Education, Training and Youth
Affairs.
Wheldall, K., & Beaman, R. (2007). MULTILIT reinforced reading: Using pause,
prompt and praise to help low-progress readers. Sydney: MULTILIT Pty
Ltd.
Wheldall, K. & Beaman, R. (2009). Wheldall, K. & Beaman, R. (in press). Effective
instruction for older low-progress readers: Meeting the needs of indigenous
students. In C. Wyatt-Smith, J. Elkins & S. Gunn (Eds.), Multiple
perspectives on difficulties in learning literacy and numeracy. New York:
Springer.
Wheldall, K., & Carter, M. (1996). Reconstructing behaviour analysis in education: A
revised behavioural interactionist perspective for special education.
Educational Psychology, 16(2), 121-141.
Wheldall, K., & Madelaine, A. (2006). The development of a passage reading test for
the frequent monitoring of performance of low-progress readers.
Australasian Journal of Special Education, 30, 72-85.
29
Wheldall, K., & Martin, B. (1977). Socio-environmental influences on the receptive
language development of young children. Midlands Association for
Linguistic Studies Journal, 2, 111-140. (Also published in Collected Original
Resources in Education, 1 (3), 3130-3159).
Wheldall, K., & Mettem, P. (1985). Behavioural peer-tutoring: Training 16-year-old
tutors to employ the ‘pause, prompt and praise’ method with 12-year-old
remedial readers. Educational Psychology, 5, 27-44.
Wheldall, K., Mittler, P., & Hobsbaum, A. (1979). Sentence Comprehension Test:
Experimental Edition. Windsor: National Foundation for Educational
Research.
Wheldall, K., Mittler, P., & Hobsbaum, A. (1987). Sentence Comprehension Test:
Revised Edition. Windsor: National Foundation for Educational Research.
30
Table 1
Means (and Standard Deviations) and the Resultant Gains on the Relevant Literacy
Variables (Raw Scores) for the Schoolwise Program in 2008
Literacy
Variable
N Pre-test
(sd)
Post-test
(sd)
Gain
(sd)
t p ES
Neale Accuracy
67
35.51
(13.01)
53.91
(15.45)
18.40
(9.10)
16.56
<.0005
1.41
Neale Comprehension
67
10.96
(5.34)
17.30
(6.37)
6.34
(4.26)
12.19
<.0005
1.19
Burt Word Reading Test
67
48.45
(13.00)
63.84
(14.90)
15.39
(7.72)
16.32
<.0005
1.18
South Australian Spelling
67
29.28
(6.30)
37.48
(7.30)
8.19
(4.35)
15.41
<.0005
1.30
WARP (wcpm)
67
81.64
(30.33)
118.84
(33.09)
37.19
(14.25)
21.36
<.0005
1.22
Martin and Pratt Non-word
67
18.82
(8.15)
30.93
(6.94)
12.10
(6.33)
15.64
<.0005
1.49
31
Table 2
Effect sizes (Cohen’s delta) comparing mean raw scores of the experimental and
waitlist control groups at post-test and also for comparisons of the mean pre and post-
intervention raw scores for the original experimental group.
Test E vs. C E Pre vs E Post
Neale Accuracy 0.98 1.07
Neale Comprehension 0.80 1.19
Burt 0.55 1.51
South Australian Spelling 0.60 1.23
WARP 0.78 1.57
Martin and Pratt Non-word 0.78 1.32
32
Figure 1.
Average weekly performance on the WARP (words read correctly per minute or
WCPM), Schoolwise Intake 1, 2008.